In many applications, light-emitting devices, for example light-emitting diodes (LEDs), are driven in a string; i.e., so that many LEDs are connected in series with each other. However, if one LED in the series-connected string of LEDs should fail, this can create an open circuit condition in which the whole LED string will then be out of operation. Time is then lost while the failed LED is detected and repaired. This is extremely inconvenient and can be dangerous when the LED string is used in applications where safety is an issue, for example, in street lighting or emergency lighting, etc. In high brightness LEDs, which are often used in these applications like street lighting, solid state lighting and emergency lighting, for example, which are intended to have low maintenance costs and a long lifetime, this situation is extremely undesirable. Other applications that are less critical, but where LED failure has an impact, are LED backlights for liquid crystal display (LCD) televisions and monitors, which are moving away from cold cathode fluorescent lamp (CCFL) backlights to LED backlights.
LEDs can be subject to failure due to an electrostatic discharge (ESD), which leads to an overvoltage condition. Operating outside its normal or ambient voltage range then causes damage to the LED. Solutions for protecting LEDs from ESD exist, which use a single diode connected in an anti-parallel configuration to the LED within the same package. This can give the LED protection against electrostatic discharges of up to 2 kV. However, although this solution offers protection to the LED in the event of an ESD, it does not prevent the loss of operation of a whole string of LEDs in the case of failure of just one LED in the string. Therefore, should the LED fail for any reason, there will still be an interruption to operation and associated costs while the failed LED in the series-connected string of LEDs is detected and replaced.